The assumption that tuberculosis spreads primarily through prolonged household exposure is being fundamentally challenged by whole-genome sequencing data from one of the world's most severe drug-resistant TB epicenters. For public health strategists and clinicians managing extensively drug-resistant strains, this recalibration of transmission routes has direct implications for how interventions are designed and where resources are directed.
A prospective molecular epidemiology study conducted in metropolitan Durban, South Africa enrolled 305 individuals diagnosed with second-line drug-resistant TB — including pre-XDR and XDR strains — between 2018 and 2022. Universal whole-genome sequencing was performed on isolates from 251 participants, with genotypic clustering defined at ≤12 single-nucleotide polymorphisms. Of the 141 participants who fell into 25 identifiable transmission clusters, 49% were linked through casual contact — geographic proximity of residences within 1 km, shared community locations, or overlapping outpatient clinic visits — compared with only 9% linked through traditional close-contact definitions such as household exposure or shared hospitalization. Multivariable modeling found living within 1 km carried an odds ratio of 17.9 for transmission linkage, making neighborhood proximity a dominant structural risk factor.
This finding sits at the intersection of two maturing fields: high-resolution pathogen genomics and spatial epidemiology. Prior TB contact-tracing frameworks were largely built around household index-case models — an architecture that this data suggests misses the majority of transmission events in dense urban settings. The implication is that community-level interventions, including improved ventilation in shared clinic waiting areas and community gathering spaces, may offer greater population-level impact than household screening alone. Key limitations include the observational design, geographic confinement to a single high-burden city, and the inherent difficulty of definitively ruling out undetected close contacts. Nevertheless, 80% enrollment and near-universal sequencing make this one of the most complete DR-TB transmission datasets assembled. The finding is potentially paradigm-shifting for urban high-burden settings globally.